Trust and Contingency Plans

Abstract

Trusting relationships are both valuable and risky. Where the risks are high and the fears of betrayal are also high, it might seem rational to try to mitigate the risks, while still enjoying the benefits of the trusting relationship, by forming a contingency plan. A contingency plan—in the sense I am interested in—involves contingent punishments for defection, which are primarily meant to encourage the trusted partner to act trustworthily. I argue, however, that such contingency plans suffer from an internal tension wherein the contingency planner both seeks and undermines a particular level (or kind) of trust. There are two problems in particular, either of which is sufficient to undermine trusting relationships: one, the planner fails to see the trusted partner as sincerely engaged in the trusting relationship, and two, the planner separates herself out from the trusting relationship by seeing her flourishing as separate from her partner’s (or, even worse, as dependent on her partner’s harm). Continency plans, then, are not just about the future; they cast a moral shadow on what we are doing now.

1. Introduction

Being in a trusting relationship exposes one to risk, and the stakes—emotional, financial, reputational—can be high. In such situations, worries that the relationship might fall through because of the trusted partner’s betrayal might tempt one to formulate a contingency plan. And so, for example, one might hope that things turn out in such-and-such a way, but on the chance that they might not, one will plan for alternative possibilities. What I would like to show is that in the context of trusting relationships, such contingency plans can be damaging to trust. When and why do contingency plans sometimes undermine trusting relationships? While there has been a burgeoning literature on the topic of trust, this particular question has been left unexplored.

Contingency plans, as “just in case” plans, can take two forms. In the first type of plan, where A and B are in a trusting relationship, B promises to A that she will φ. Then A, in order to motivate B to keep her promise just in case she cannot sufficiently motivate herself, announces a hypothetical punishment of the form, “If you don’t φ, then I will ψ.” Such plans are based on a prediction of how the trusted partner might fail, and are primarily meant to encourage that partner to act consistently with the demands of the trusting relationship. Let us call this a ‘paternalistic’ contingency plan.

The second type of contingency plan, which is perhaps the more familiar type, is the Plan B. The ‘Plan B’ is a kind of second-best planning, where the planner makes a backup plan in case her first-choice plan does not pan out. For example, Molly really wants to be a philosophy professor but also trains as a lawyer in case academia does not work out. In the context of trusting relationships, Plan Bs have the function of protecting oneself (or minimizing harm to oneself), perhaps by deflecting harm onto the person who defects first should the trusting relationship fail. ¹ Both types of contingency plans— paternalistic plans and Plan Bs—share the form, “If you breach our trust, I will ψ,” but, as I will show, they undermine trust in different ways.

I begin in section 2 by examining how contingency plans affect the development or maintenance of trust in the context of interpersonal relationships. In section 3, I move to civic relationships (relationships we have as comembers of a political community), and consider how permissive gun laws affect civic trust. Although different in many important respects, I hope to show that contingency plans at both the personal and civic levels share a structural similarity, such that they undermine trust. Finally, in section 4, given the problems with contingency plans, I draw some initial conclusions about what moral reasoning in the context of trusting relationships might look like.

2. Contingency plans and interpersonal trust

To get started on the problem of contingency plans, consider the following example of a paternalistic plan:

Dinner Plans: Abby and Andy are good friends, and are making plans for dinner at a nice restaurant. Feeling frustrated with Andy’s chronic lateness, Abby asks for and receives Andy’s promise that he will be punctual. Abby, however, is not entirely satisfied with the promise, and in an effort to incentivize Andy’s being on time, says to him with a friendly smile, “If you’re late, then dinner’s on you.” Andy is hurt by Abby’s remark.

How should we think about Abby’s paternalistic contingency plan, and about Andy’s hurt feelings? The answer will depend on why Abby feels that the promise is inadequate and a contingency plan is necessary. In this case, it seems that Abby is motivated by something like the following: Abby sees herself as helping Andy to keep his dinner promise and to be a good friend. She sees herself as helping him to do something he already wants to do, which is to be respectful of Abby’s time and to be a good friend. Given Andy’s history of lateness, Abby believes there is a good chance of Andy’s being late again in the future. But with a contingency plan in place, Abby predicts that the chances of Andy’s being late will go down.

If this is what is motivating Abby, why might Andy be justified in feeling hurt? The problem is that Abby is failing to see Andy as living up to his obligations as a good friend, and her contingency plan is keyed to her prediction of how he might fail. And she is using her intimate knowledge of Andy to try to prevent him from failing her. For example, the punishment she chooses is paying for an expensive dinner because the cost is sufficiently high that it would be a hardship for Andy. When Abby thinks about Andy in these ways, she is treating Andy as a variable to be managed, rather than as an agent who acts for reasons, like choosing to keep his promise. Andy might feel offended or disrespected in this case because he sees himself as a person who will keep his promise, and Abby refuses to see him that way.

Consider now another example of a contingency plan, this time a Plan B:

Prenup Demand: Ben and Bianca decide to get married. They take marriage to be a lifelong commitment to support and love each other, which they believe requires fidelity. A short while after they make their decision to get married, Bianca presents Ben with a prenuptial agreement that specifies that in case Ben is unfaithful in the marriage, he will have to pay Bianca $1M in divorce. Ben is hurt. ²

How should we think about Bianca’s contingency plan, and Ben’s hurt feelings? Again, the answer depends on Bianca’s motivation. Perhaps, like Abby, Bianca sees herself as helping Ben to keep his promise. Then the contingency plan is functioning as a paternalistic plan, and Ben might be upset for the same reasons as Andy.

But perhaps Bianca is being motivated by fear or worry about Ben’s infidelity, and the contingency plan is a way for her to mitigate her losses should the feared thing happen. In this case, while her prenup demand might be functioning primarily as a contingency plan, it is also functioning secondarily as a Plan B. Bianca wants to make sure she will come out on top, or at least properly compensated for the time and effort she puts into the marriage. In this case, the prenup is something like an insurance plan: Bianca hopes the bad thing will not happen but wants to be prepared in case it does.

Why might Ben be justified in feeling hurt in this case? I see two different possibilities. First, if Bianca’s plan is to come out on top, the goal of her Plan B is to shift the cost of the disintegration of the relationship onto the party who betrays first. It is a way to try and make sure that if the bad thing happens, Bianca can deflect as much of the harm away from herself and onto Ben. In this case, Ben might be hurt that Bianca sees him as a potential target of harm, or at least seeks to promote her own well-being at his expense.

Second, if Bianca’s plan is to make sure that she is properly compensated for the time and effort she puts into the marriage, Ben might be hurt that Bianca is separating herself out from their relationship. She sees her future well-being, or at least one avenue of her future well-being, as separate from (and in some sense opposed to) Ben’s well-being.

In his discussion of trustworthiness, Russell Hardin considers the rise of the prenuptial agreement in a divorce-tolerant society. Against the backdrop of a society like ours that is generally accepting of divorce and has high divorce rates, he asks us to imagine two people who exchange wedding vows and promise fidelity. Should they believe each other? Hardin answers his own question thusly (1996, 36–37):

Perhaps. But perhaps only in the sense that we really are committed at this moment. … For commonsense epistemological reasons, we individually have to live with the larger society’s coordination on a particular pattern of expectations. For example, I cannot reasonably expect you to be dramatically different in your long-run commitments than others are.

Because “infidelity and opportunities for it are rampant,” Hardin goes on, those who exchange wedding vows “may be unable to trust one another with anything short of an act of faith” (1996, 37).

I think that reasoning like Hardin’s threatens to make marriage, as he conceives of it (i.e., a lifelong commitment based on trust), impossible. When A and B take themselves to be in this special kind of relationship, governed by fidelity and trust, and each reasons from the fact that spouses are sometimes unfaithful to the necessity of a prenup, they make it difficult (and perhaps impossible) to actually relate to each other as marriage partners. By preparing right now for the end of their relationship, each is taking the fact of their special friendship as giving her reasons in the wrong way. Namely, she is taking her friendship as a token of a type of relationship that is sometimes betrayed and seeing that as giving her reason to act right now as if she might be betrayed.

It might be objected in defense of both Abby and Bianca that there is no moral problem with their contingency plans—neither can their partners reasonably be offended, nor are their contingency plans incompatible with a trusting relationship. We should understand Abby, rather than treating Andy as a variable to be managed instead of as an agent (as I have argued), as actually just giving him an additional reason to do the right thing. With Abby’s contingency plan in place, Andy not only has the reasons he already had to be on time, he also has the additional reason of avoiding the cost of an expensive meal for two. And with respect to Bianca, if she is reasoning along the same lines as Hardin, then by considering facts such as her partner’s past infidelity and social norms concerning divorce, she is merely being prudent in considering facts that are relevant to her decision to marry.

To see why Abby’s and Bianca’s contingency plans are incompatible with trust (and so why their partners might be offended), we can look at how they are both engaged in probabilistic reasoning. Abby cannot bring herself to take Andy at his word that he will be on time because Andy has often been late. And Bianca cannot bring herself to believe Ben’s promise that he will be faithful because Ben has been unfaithful in the past (and perhaps also because of the divorce statistics in her community). Abby and Bianca are treating Andy’s and Ben’s promises that they will φ as merely another piece of evidence that bears on the question of whether Andy and Ben will in fact φ. What Abby and Bianca should be doing, however, is treating the promises as what Berislav Marušić calls reasons of trust.

When considered as a reason of trust, “[S]omeone’s promising us to φ consists in an offer of an answer to the question of whether she will φ … If we accept the other’s offer, we trust her” (Marušić 2015, 195). If Andy and Ben are sincere in their promises, then they have decided for themselves that they will φ, where φ-ing is up to them and in their control (i.e., being on time and being faithful, respectively). Their sincerity points to an important asymmetry between agents and outside observers. If Andy and Ben reason as outside observers, they treat the question of whether they will φ as a theoretical question and rely on evidential reasons—and they may come to the conclusion that they will not follow through on their promises. But, as Marušić explains, to reason in this way about one’s own actions is to fail to take responsibility for one’s agency (2015, 137–38). Instead of predicting what they might or might not do, Andy and Ben must decide what to do. They must treat the question of whether they will φ as a practical one, not as a theoretical one, and so rely on practical reasons, not evidential ones. And it is these practical reasons that will ground their practical beliefs about what they will do—namely keep their promises. ³

So then the problem with Abby’s and Bianca’s reasoning is that the nature of their trusting relationship requires them to be partial to Andy and Ben, and to take them at their word. ⁴ And we can see more clearly now how trusting is not compatible with doubts about the other’s ability or willingness to act as trusted—doubt is based on evidential beliefs, while trust is based on practical beliefs and accepting another’s word as settling the question of whether they will φ.

We might well wonder at this point what it is that people who are in trusting relationships trust each other to do. The scope of trust will be particular to each relationship based on the partners’ goals, practices, and expectations. Often, it will not be possible to enumerate, in advance, all of the particular ways in which partners trust each other, or all of the particular things partners trust each other to do. This is especially true in relationships like marriages or friendships, which unfurl over long and indefinite periods of time, where the scope of trust will necessarily be open-ended. But even then, partners can make their general expectations clear, for example, when they promise to love, honor, and cherish each other all the days of their lives.

Of course, there are times when it is not at all clear whether partners should trust each other, or whether the scope of trust should be expanded, and I will have more to say about those situations in section 4. But the cases at hand—Dinner Plans and Prenup Demand—are simpler, because of the promises involved. Andy’s promise to be punctual is based on his consideration for Abby’s wishes, understanding how important it is to her that he keep his promise, and being respectful of her time and feelings. And Ben’s promise to be faithful is partly constitutive of the marriage relationship as he and Bianca conceive of it. Andy and Ben have each given their word that they will φ, where φ-ing is either an important (or constitutive) part of their relationship, and/or evidences some value or values that are important to the relationship.

We are now in a position to see more clearly the tension produced by contingency plans like Abby’s and Bianca’s. On the one hand, Bianca and Ben (and Abby and Andy) understand their relationship as governed by trust. On the other hand, neither Bianca nor Abby can bring themselves to believe that their partner will act trustworthily, and so each creates a contingency plan meant either to help their partner act in trustworthy ways or, in Bianca’s case, to protect herself in case of Andy’s betrayal. Bianca, or Abby, is then in an untenable position of (1) seeking a certain level or kind of trust in her relationship, but (2) at the same time undermining it by (a) failing to see her trusted partner as living up to his commitment and/or (b) separating herself out from their relationship. In constructing a contingency plan, Bianca and Abby are themselves failing to realize the trusting relationship that each seeks, such that even if Ben is faithful and Andy is on time, each has already damaged the trusting relationship that she desires.

To get clearer on the ways in which Dinner Plans and Prenup Demand are problematic as contingency plans, compare them to the following case:

Prenup Negotiation: Pete and Pat decide to get married. They will promise to be faithful their whole lives. But just in case things do not work out, they would like to make clear, together, how to separate their property upon divorce.

Notice first that this case seems to lack the two key problematic features of contingency plans: (1) their prenup negotiations are not meant to keep each other in the marriage, but rather contemplate a mutual exit; and (2) the separation of property is not meant as a hypothetical punishment, but as a mutually desired outcome. And so Prenup Negotiation seems like it should not share in the same kinds of problems as paternalistic plans or Plan Bs.

Whether Prenup Negotiation is problematic for trust will depend, I think, on what Pete and Pat take the marriage promise to involve. Do each of them vow, “I promise to love and honor you all the days of my life,” or “I promise to try my best to love and honor you all the days of my life”? The former formulation makes the prenup negotiation harder to square with trust, for then Pete and Pat should take each other’s promise to mean that each will do the thing promised. But I do not want to resolve this case one way or the other. I think it’s enough to say, for my purposes, that if Prenup Negotiation is inconsistent with trust, it could be for this reason.

To summarize, a contingency plan is a “just in case” plan premised on the trusted partner not performing in the expected, trustworthy way. A paternalistic plan is meant to keep the other person in the relationship by compelling or encouraging them to act trustworthily via a system of hypothetical punishments; and a Plan B is meant to protect the faithful person in case of defection. Such plans are problematic because they fail to see the other as living up to their obligations, and/or one or both parties separate themselves out from the relationship.

3. Contingency plans and civic trust

It might seem from the choice of the examples discussed above that trust is something built up over time between two or more particular persons. But, in fact, trust is a very ordinary part of our relationships with strangers. ⁵ We trust strangers all the time, otherwise a society such as ours, where our daily activities are marked by interactions with people we do not know, would not be possible. For example, when I am riding in an elevator with strangers, I do not worry that they will try to steal my lunch. That I trust them not to steal my lunch need not be an occurrent thought. Rather, such trust is part of the background condition that makes it possible for me to live my life around and with other people I do not know.

What the appropriate standard of trust is between strangers of a particular community will vary greatly and is sensitive to innumerable factors. There are many particular standards of trust and learning what the appropriate standard consists in is part of the process of enculturation. And so, for example, as a visitor traveling with my baby to Copenhagen, I could not leave my baby unattended outside on the curb while I ducked into a café to grab a quick lunch, as seems to be the Danish norm. Because I am not part of their community, I have not been trained to see the world in ways that support the level or type of trust that the residents of Copenhagen have in each other, and so the risks involved in leaving my baby unattended are intolerable for me. Similarly, tourists from Copenhagen visiting Manhattan should not leave their baby unattended outside on the curb while they grab a quick lunch. ⁶

Now consider the following example:

Bus Riding: Betty is riding on the bus. Her mind starts to wander, and she has an idle thought because of a movie she saw last night: What if some people on the bus were out to get me? She starts to think up a hypothetical plan: If some people on the bus were out to get me, what would I do? She considers this for a moment, and starts to scope out the nearest exit, and makes sure it is clear; she notices that the umbrella, which she had been holding onto all along, could be used as a weapon; and her eye keeps wandering back to the door to make sure the path remains clear.

It is unclear at what point Betty’s idle thought becomes a hypothetical plan and when it becomes a contingency plan (if ever). But what is interesting about Betty’s stream of thoughts is this: it shapes the framework within she interacts with others on the bus. If she is looking for escape routes, and the next person who gets on the bus stands in front of her, she will see that person (in some way) as an impediment to her escape route, and therefore as a danger to her, even if only a hypothetical danger. What we see from this example is that contingency plans seep into what someone is doing now, even if on the outside it does not look like anything is changing (e.g., maybe Betty already had her umbrella in her hand, and she was already sitting next to the exit).

Now what about explicit contingency plans that involve violence, again in the public sphere? Take a particularly American example of a violent contingency plan—permitting ordinary citizens to carry guns. One justification often appealed to is that carrying guns makes us all safer. If more citizens carry guns, the bad guys will be under increased risk that they themselves will be shot if they try to use violence against law-abiding citizens, and so they will be less likely to attack law-abiding citizens. This is a paternalistic contingency plan. Compare this rationale to that of another kind of gun buyer; call him ‘X.’ X buys a gun because he has some fear or expectation that civil society will break down, and that he will find himself in a state-of-nature-like situation. X’s contingency plan is of the Plan B variety. To understand what might motivate either line of reasoning, I think it is useful to look at the Hobbesian notion of anticipation.

According to Hobbes, the most “reasonable” strategy when living in a state of nature, and so in a condition of mutual fear, is “Anticipation; that is, by force, or wiles, to master the persons of all men he can, so long, till he see no other power great enough to endanger him” (1991a, 87–88). Hobbesian anticipation is attacking first and killing others before they can kill you (or coercing others into submitting to you so that you can put their power at your disposal), because you expect that others may try to attack or kill you. ⁷ The more general idea is that a person living in a state of nature, out of her fear of other people, must use violence or be prepared to use violence to preserve her own life.

So, what happens when civilians arm themselves against each other? Even if citizens who carry guns never fire a single shot, the nature of their relationship to their co-citizens changes. When we are out and about in public spaces, we all have a certain perimeter around us, marking our “personal space,” which we prefer that others not cross. The amount of personal space we need is context dependent—personal space on a crowded subway is different from personal space on an empty subway. When this perimeter is crossed, it can make us feel uncomfortable, nervous, or unsafe.

In order for guns to be effective at protecting our perimeter (and so our bodies and our sense of well-being), or effective in the way that supporters of permissive gun laws imagine, the attacker has to not already be in your personal space. Otherwise what ensues will be like a scene from a movie, with two people struggling over one gun. This means that in order for a gun to be effective in the way the NRA supporter imagines, we need to dramatically enlarge the personal space around ourselves in order to feel safe. We have to defend a lot more personal space. Then, like Betty on the bus, we have changed the framework within which we interact with each other. We see each other not as people going about their business, not as people who we might ask for help if we are lost or if we slip and fall, but as people coming in and out of each others’ territory—as potential threats. And these kinds of thoughts, once they become pervasive enough, destabilize civic trust. ⁸

We see the same kind of tension here in the NRA supporter’s contingency plan as we did in Abby’s and Bianca’s contingency plans. The NRA supporter seeks to live in a society where people are free to live their lives, free from the threat of assault or battery. ⁹ But in order to be free to live their lives in this way, people must be unburdened from having to enforce their own rights against assault or battery; such a burden would be exhausting (people would not have time or energy left to devote their attention to pursuing their vision of the good life) and it would be impossible for people to actually enforce their rights in this way. ¹⁰ Citizens need habits of obedience or moral beliefs that following the law is the right thing to do. And so underlying the kind of society that the NRA supporter seeks is the assumption of a civil society.

The tension with the NRA supporter’s position should be clearer now. At the same time that he seeks a society based on civic trust, he himself, in constructing a paternalistic plan, fails to see others as living up to their commitment. Because effective use of his gun requires him to monitor and maintain a large perimeter around himself, he might see anyone—not just the bad guys—as potentially failing in their commitment. And/or, in constructing a Plan B, he is separating himself out from his co-citizens by seeing his own flourishing, qua member of civil society, as separate from others’, or seeing others as potential target of harm. He does not want to live in a state of nature, but by regarding others in anticipation, he is relating to them as if he is in a kind of state of nature. Once enough people start acting like the NRA supporter, civil society becomes weakened or broken down because it requires (enough of) us to comport ourselves to an ideal of civic trust.

4. Trust and moral reasoning

Having looked at the ways in which contingency plans interfere with the development or sustaining of trusting relationships, we might be in a position to draw some initial conclusions about what moral reasoning in the context of trusting relationships should look like.

First, when reasoning about what to do in the context of a trusting relationship, our reasoning should not be strategic, relying primarily on self-regarding assessments about risks and benefits. In trusting relationships, our thoughts should not be about pursuing an individual goal or separating ourselves out from the relationship. To reason ‘strategically,’ as I will use the term, is to instantiate a relationship where one seeks to use the other in order to further one’s own plans or goals, and is especially objectionable when done in ways that are inconsistent with the parties’ mutual understanding of their obligations to each other.

That a friend ought to refrain from reasoning strategically does not mean that she is unable, in the sense of literally incapable, of making predictive evaluations. Indeed, knowing what your friend is likely to choose between chocolate cake and carrot cake can be a sign of your intimacy. Rather, what it means that a friend refrains from reasoning strategically is that she does not set out with some personal or private goal, and then act on the basis of that predictive evaluation in order to “manage” her friend as a variable who might help or hinder the attainment of that personal end.

Second, I recognize my trusted partner as someone who is morally competent, and whose moral competence is directed at me in particular (qua friend, or qua co-member of a political community). Let me call this ‘goodwill’ for short. I assume my trusted partner’s goodwill, and because I have goodwill toward my trusted partner, our relationship is marked by reciprocal goodwill. ¹¹

We should not understand goodwill as merely good feeling or benevolence ¹²—a sturdier anchor is needed to account for the stability of trusting relationships. When I decide to trust, I choose to vest the trusted partner with some decision-making authority in my life on the belief that my own interests feature appropriately. As Barbara Herman argues, acting for the right sorts of reasons is constitutive of a certain kind of relation. The fact that someone is my friend means

that I have reasons for action of a certain sort. Having these relationships is to have these reasons … They are reasons such that acting on them (and not on other reasons that can produce the same outcome) is important to maintaining the relationships that generate them. (1991, 780)

Since my belief in my trusted partner’s goodwill is the basis for my vulnerability to her, my belief in her goodwill must involve a judgment about her moral competence—that she will do the right thing, at the right time, for the right reasons, where my flourishing is one of her central reasons. Because trust entails moral deference, it should be seen as a form of respect.

My expectations concerning my friend do not have to do primarily with whether or not she will perform some particular action. This openness makes it the case that it would not be appropriate for me to feel betrayed just in virtue of the fact that my trusted partner, in some particular circumstance, undertook some action different from what I would have liked, or what I would have chosen. (But if she exercises judgment badly, that, of course, is another matter.)

Because the appropriate level of goodwill—understood as moral competence aimed at a person (or class of persons) in particular—is highly context dependent, there cannot be a single standard by which we can measure goodwill. ¹³ For me to have the appropriate kind of goodwill toward you is for you to matter in the way that you should in my life. ¹⁴ And so goodwill between a parent and child will look different from goodwill between two casual friends, which will in turn look different from goodwill between two intimate friends.

Third, our openness to our trusted partner exercising their moral judgment in our lives helps to explain why we must be open to the possibility of betrayal. One of the characteristics of trust is that it is resistant to evidence of betrayal in ways that predictions are not. Karen Jones uses the example of Iago and Othello (1996, 11–13). If Othello had trusted Desdemona, he would not have been so easily swayed by Iago’s insinuations about Desdemona’s infidelity. To put it another way, even if the doubts that Iago raised had been reasonable (in some sense), Othello should have trusted Desdemona, where that would have involved, at a minimum, withholding judgment about Desdemona’s guilt and seeking an alternative explanation.

But there is another way in which trust makes us vulnerable to betrayal. For Othello, when certain things were brought to his attention, he interpreted them as evidence of Desdemona’s infidelity. But when we trust, certain things do not even show up for us as evidence. There is a difference between what Othello does, namely interpreting evidence, and not even seeing that something could be interpreted as evidence. There is simply no recognition that something needs to be interpreted as either fidelity or infidelity.

Consider the following scenario: Desdemona* is home late from work yet again. Othello* could think to himself, “It sure is unusual for Desdmona* to be home so late so often. It’s probably nothing.” Or he could think to himself, “Poor Desdmona*. She’s been working so hard lately.” In the first thought, Othello* interprets Desdemona*’s lateness charitably—i.e., he considers and rejects the possibility that something illicit is going on. In the second thought, Othello* does not interpret Desdemona*’s lateness as anything; he only recognizes the fact of her lateness as reason for sympathy. So the way our deliberative field is constrained in trusting relationships potentially blinds us to evidence of betrayal twice over. ¹⁵

It might be objected at this point that on my account of trust, and especially given the necessity of vulnerability, it will never be permissible to reevaluate trusting relationships—e.g., to have thoughts or questions about a trusted partner’s trustworthiness—on pain of violating trust. ¹⁶ It seems that trust will not permit me to recognize that there is a pattern in the way my partner lets me down, or that my partner has certain faults or weaknesses.

This is a worry that is not easily dealt with. But to start, it is important to remember that the ways in which it is appropriate for me to make myself vulnerable to you will depend on the goodwill involved. Both too much vulnerability and too little can be damaging to trusting relationships; with too little, it will be difficult to establish a trusting relationship because there might not be enough room for the partner to exercise judgment and for trust to operate, and with too much, the relationship can become burdensome for the trusted partner or dangerous for the one who trusts.

And any account of trust will have to be sensitive to the fact that trust is not all-or-nothing—i.e., not only do we trust different people to different degrees, but even within the context of a trusting relationship, the boundaries of trust can be dynamic. In general, certain kinds of thoughts (e.g., the likelihood that a friend will fail to live up to her obligations) ought to be excluded from our deliberative field. But in new, or unfamiliar, or particularly risky terrain, considering the possibility of what a friend might do—whether she will fail or succeed in living up to our expectations—is not necessarily inconsistent with trust. And that is because the novel situation is outside the scope of our mutual understanding of our obligations to each other. In the face of such uncertainty, however, a friend can choose to trust.

It is true, of course, that in general we cannot bring ourselves to trust at will, any more than I can bring myself to believe that the earth is flat. ¹⁷ But trust goes further than I have reason right now to believe, and so the fact that I do not have inductive grounds to believe that a friend will φ does not mean that I do not have grounds to trust that she will φ. ¹⁸ For in trusting, I hold my hand out to the future, because I believe my friend will meet me there to grab it. ¹⁹ If I cannot genuinely bring myself to believe that my friend will meet me in this way, and yet choose to act as if she will, then I am not so much acting on the basis of trust as I am taking a leap of faith. ²⁰

5. Conclusion

Trusting relationships are valuable, and they are also risky. Contingency plans might seem like a promising way to navigate this terrain by allowing the planner to foreclose or mitigate the chances of betrayal, while still supporting the trusting relationship. What is interesting about the contingency plans discussed in this paper is that they are all tailored preparations for breaches of the trust in a relationship based on predictions of the trusted partner’s failure or betrayal, with a possible secondary aim of coming out on top, or at least flourishing separately from the trusted partner. Again, none of what I have argued in this paper is to say that exiting trusting relationships is bad, or that planning together to exit a relationship is necessarily bad.

Rather, I have shown that there are two problems with such contingency plans, either of which is sufficient to undermine the kind of trusting relationship sought: first, the planner fails to see the trusted partner as living up to her commitment; and second, the planner separates herself out from the relationship by seeing her flourishing as separate from the flourishing of the partner (or worse, by seeing her flourishing as possible only through the other’s harm). I have tried to show that contingency plans suffer from an internal tension: the contingency planner both desires a certain kind of trusting relationship and undermines the relationship she desires by planning for its potential collapse. The contingency plan corrupts the planner’s moral reasoning in that she cannot reason appropriately given the kind of trusting relationship she seeks. Contingency plans, then, are not something that merely happen in the future; rather, they cast a moral shadow that changes the nature of what we are doing now.